Measure XRD sample preparation height to correct peak shifts

[rekumar]

Automated XRD sample prep can sometimes lead to piles of powder that are taller than we want. This leads to a positive shift in diffracting angles. If we can measure the height of this pile, hopefully we can correct our diffraction pattern.

[rekumar]

I have a meeting with SICK sensors to discuss distance sensor options. I will ask about ultrasonic vs optical sensors, particularly interested in sensitivity to material properties.

If we can identify the right sensor, we could position it to measure the height of powder on the XRD holder during sample prep. This could be:

• at an angle to provide a clear path for the robot arm.
• directly above the xrd sample holder, but then we need to move it in/out of position with a linear actuator

[bernardusrendy]

Yuxing measured MgO XRD with different heights (minimum height by human, higher by the robot arm, and even more heightened by human). Bernard and Yuxing tried refining it to get different sample displacements. This is to do a sanity check. Results are coming soon.

[njszym]

Yuxing had the robot prepare the four duplicate samples again, now with 20% wt. NaCl included. By refining the patterns from these samples, he was able to identify the peak shifts and correct for them. On the corrected patterns, autoXRD gave much more consistent predictions.

Moving forward: more tests need to be done to determine whether the shifts are uniform. If so, we can apply some average shift for each sample (without requiring the addition of some internal reference). If this doesn't work, perhaps we can add a bit of the known reference to each sample.

[bernardusrendy]

Need to rethink if this is still necessary given the refinement we could already do?